[hpsdr] Direction Finder using Dual-Mercury Diversity Operation

[Joe Martin K5SO]

Hi John,

You bring up a good point.

Using a common 122.88 MHz clock for all Mercury boards does allow  
stable phasing as shown by long-term (hours) nulling of a signal  
generator signal into the Mercury boards.  The null position doesn't  
change upon exiting and re-starting PSDR but DOES change to a new  
stable phase relationship when the Atlas board is powered down and  
restarted.

As you note, each time the boards are re-started the relative phase  
will be different but will remain constant as long as power is not  
removed from Atlas.  As each Mercury board is constructed internally  
the same way and each is working from exactly the same instruction set  
passed to it (them) from Metis/Ozy/Magister the various delays are at  
least constant for the Mercury boards after the ADCs are started up.   
This does enable deep, stable nulls to be observed with dual Mercury  
boards in a laboratory test setup.

Now, what I should find out is whether, once the direction is  
calibrated in a multiple Mercury setup, the calibration changes every  
time Atlas is re-started?  I don't really know.  I'll have to check  
that out.

73,  Joe K5SO


On Sep 22, 2011, at 2:06 PM, John Ackermann N8UR wrote:

> ***** High Performance Software Defined Radio Discussion List *****
>
> Just FYI -- I've seen some references on the GnuRadio list that  
> indicate just locking clocks across multiple hardware units may not  
> be enough for phase-critical applications.  Instruction sequencing  
> and various other delays mean that the initial phase of the two  
> systems is likely to be arbitrary -- each time the code runs the  
> relative phase will be different (but will remain constant during  
> that run).
>
> I'm *not* knowledgeable about this; just passing along gossip that  
> might be relevant.
>
> John
> ----


 1316726103.0